Predicting Kurdish EFL University Learners' Oral Reading Fluency Using Support Vector Machine

Investigating learners’ English Oral Reading Fluency (ORF) in the contexts where is used as a foreign language (EFL) or second (ESL) has recently become trending subject. This study was carried out to predict ORF of 100 Kurdish EFL university students department, college Basic Education, Duhok, Iraq 2020 during covid-19 using support vector machine (SVM) technique. technique one supervised learning techniques, and it considered most powerful algorithm terms high accuracy; therefore, employed this study. Participants’ measured by two experienced human raters four dimensions Multidimensional Scale (MDFS) including expression & volume, phrasing, smoothness, pacing, which were input variables an output. Six kernels SVM prediction process. The results indicated that highest accuracy testing result obtained on use Linear kernel with value 96.2%. Confusion matrix utilized assess outcomes data classification. precision, recall, F1-score for linear their values same all performance metrics 96.1%. Accordingly, can be concluded considerably accurate predicting oral reading fluency.